Microlens arrays provide optical versatility in a miniature package for imaging applications. Traditionally, a microlens is defined as a lens with a diameter less than one millimeter; however, a lens having a diameter as large as five millimeters or more has sometimes also been considered a microlens.
There are many conventional methods for manufacturing microlenses. For example, one commonly used technique for manufacturing microlenses begins by coating a substrate with a selected photoresist, exposing the photoresist coated substrate to radiation through a mask, or alternatively, subjecting the photoresist to gray scale laser exposure. Upon heating the substrate, the exposed photoresist melts and surface tension pulls the material into the form of convex lenses. The depth of the photoresist determines the focal length of the lens.
Another method for the manufacture of microlenses is to use ion exchange. In this method, ions are diffused into a glass rod to give a radial refractive index distribution. The index of refraction is highest in the center of the lens and decreases quadratically as a function of radial distance from the central axis. Microlenses made using the ion exchange method are used to collimate light from fibers as, for example, in telecommunication applications.
In general for many applications, microlens arrays are preferred over discrete microlenses. As an example, one manufacturing process for the production of glass microlens arrays generally involves reactive ion etching (RIE) of fused silica. In general, it is very difficult to meet all the requirements of microlens arrays using RIE. The RIE technology involves many steps before the final product can be produced and thus the yield is typically poor and the products are costly.
As another example, compression molding of optical quality glass to form microlens arrays is also well known. This method includes compressing optical element preforms, generally known as gobs, at high temperatures to form a glass lens element. In the compression molding process, a gob is inserted into a mold cavity. The mold resides within an oxygen-free chamber during the molding process. The gob is generally placed on the lower mold and heated above the glass transition temperature and near the glass softening point. The upper mold is then brought in contact with the gob and pressure is applied to conform the gob to the shape of the mold cavity. After cooling, the lens is removed from the mold.
Unfortunately, compression molding an array of microlenses using one or more preforms is subject to many difficulties, which may include alignment of mechanical and optical axes of each lens element with respect to a common axis and location of each lens element with respect to a reference point in the array. Furthermore, it is extremely difficult to machine convex aspheric mold cavities using conventional techniques if the microlens diameter is less than 1 mm.
As another example, microlens arrays are often formed on the top surface of silicon chips, either for light-sensitive (e.g., CCDs) or light-emitting (e.g., micro-display devices) applications. A planarization layer is first formed over the silicon substrate. A color filter layer is next formed over the planarization layer with sub-pixel areas properly aligned with active devices in the silicon substrate. Another planarization layer is generally formed over the color filter layer and, finally a photoresist material is deposited over the second planarization layer. Conventional lithographic techniques are then utilized to form rectangular patterns in the photoresist. After exposure, a development step removes the photoresist in the exposed areas leaving the central island regions over the pixel-active areas transparent. Development and sometimes etching, removes the photoresist material between these central regions and forms trenches in the photoresist area separating the islands of photoresist now defining the individual microlens sites. A deep plasma etch into the silicon substrate next removes all layers above the substrate. Photoresist is then stripped and the devices are hard-baked to reflow the micro lenses into the proper optical form by controlling time and temperature.
Although there are many conventional methods for producing microlenses and microlens arrays, these conventional techniques may involve difficult or expensive manufacturing steps or fail to meet certain design requirements, such as for example in terms of viewing angle, brightness, uniformity, or contrast. Consequently, there is a need for an improved microlens array.